These slides presents a method for creating coherently animated line drawings that include strong abstraction and stylization effects. These effects are achieved with active strokes: 2D contours that approximate and track the lines of an animated 3D scene. Active strokes perform two functions: they connect and smooth unorganized line samples, and they carry coherent parameterization to support stylized rendering. Line samples are approximated and tracked using active contours ("snakes") that automatically update their arrangment and topology to match the animation. Parameterization is maintained by brush paths that follow the snakes but are independent, permitting substantial shape abstraction without compromising fidelity in tracking. This approach renders complex models in a wide range of styles at interactive rates, making it suitable for applications like games and interactive illustrations.

1. Pierre Bénard1,2 Jingwan Lu3 Forrester Cole4
Adam Finkelstein3 Joëlle Thollot1,2
1Grenoble University, LJK 2INRIA 3Princeton University 4MIT CSAIL

2. 2

3. 3

4. • Image space buffers • Object space line
processing extraction
Comprehensible rendering of 3-D shapes, Real-Time Nonphotorealistic Rendering,
Saito and Takahashi, SIGGRAPH 1990 Markosian et al., SIGGRAPH 1997
4

5.  Simple and fast
 Natural coherence and LoD
 Restricted stylization effects
Implicit Brushes for stylized line-based rendering,
Vergne et al., CGF 2011
5

6.  Wide range of styles
 Computationally expensive
 Complex LoD
 No natural coherence
A Procedural Approach to Style
for NPR Line Drawing from 3D models,
Grabli et al., TOG 2010
6

7. Line texture
l
0 Brush path
7

8. Fixed Line View-Dependent Lines
 Creases  Silhouettes
 Ridges and valleys  Suggestive contours
 Apparent ridges
Parameterization = arc-length Parameterization = ?
8

9. • Flatness
 Remain linear in image space
• Motion coherence
 Evolve according to the motion of the object
• Temporal continuity
 Adapt to the topological events
9

10.  Optimization between
2D and 3D
 Multiple brush strokes
per line
 Dependent on the
input connectivity
Coherent Stylized Silhouettes,
Kalnins et al., SIGGRAPH 2003
10

11. 11

12. Kalnins et al., Coherent Stylized Silhouettes, SIGGRAPH 2003
12

13.  2D Infinite zoom:
Self-Similar Line
Artmap (SLAM)
 Dependent on the
input connectivity
 One brush stroke
per line
Self-Similar Texture for Coherent Line Stylization,
Bénard et al. NPAR 2010
13

14. Spatio-Temporal Analysis for
 Ease propagation Parameterizing Animated Lines,
of parameterization Buchholz et al., NPAR 2011
with CSS
 Input connectivity
 Optimization over the
Snaxels on a Plane,
entire animation
Kevin Karsch and John C. Hart,  Offline computation
NPAR 2011
14

15. • Image space active contours [Kass et al. 1988]
 Shape
 Topology
 Density
• Brush paths
 Coherent parameterization
 Shape abstraction
• Interactive frame rates
15

16. 3D scene Feature samples
Image space
lines extraction
Line drawing
Snakes
Brush Paths
Tracking
Geometry
Advection Relaxation
Parameterization
Vectorization
Stylization
Coverage Connectivity
16

17. • Feature samples extracted in image space
 2D position
 Local tangent
 2D velocity
17

18. 3D scene Feature samples
Image space
lines extraction
Line drawing
Snakes
Brush Paths
Tracking
Geometry
Advection Relaxation
Parameterization
Vectorization
Stylization
Coverage Connectivity
18

19. Vectorization Tracking
 Coherence  continuous evolution across frames
 Accuracy  faithfully representation of shape
 Coverage  level of detail
 Simplicity  simple topology
 Length  stylization freedom
19

20. • Color regions tracking
in videos
 Off-line computation
 User corrections
SnakeToonz, Agrawala
NPAR 2002
Keyframe-Based Tracking for
Rotoscoping and Animation,
Agrawala et al. SIGGRAPH 2004
20

21. • Advection + Relaxation
Frame f Frame f+1
21

22. • Reprojection
similar to Bousseau et al. 2007 and Lu et al. 2010
Image f
f+1
22

23. • Minimize the energy:
Internal External
– Continuity Attraction by the
– Smoothness features samples
• Semi-explicit Euler scheme
23

24. • Grow / shrink
 re-sampling (similar to Delingette et al. 2000)
• Contour shrinking
 mass-spring forces between the vertices
 almost constant length if no tangential
external force is applied
24

25. • Advection + Relaxation
25

26. • Local vectorization operators
• Applied sequentially in a greedy fashion
26

27. 27

28. 28

29. 3D scene Feature samples
Image space
lines extraction
Line drawing
Snakes
Brush Paths
Tracking
Geometry
Advection Relaxation
Parameterization
Vectorization
Stylization
Coverage Connectivity
29

30. • Linear image space parameterization:
with = slope
= phase
= arc-length
• Evolve according to the motion and topology
of the snakes
 parameterization at each vertex
30

31. • Propagation at each vertex
 Parameterization stored between two frames
31

32. • Propagation at each vertex
• Linearization (in the least-square sense)
parameterization
arc-length
32

33. • Propagation at each vertex
• Vectorization events
Linearization
 Split
 Extend propagated directly
 Trim
 Merge: mechanism to avoid parameterization
discontinuities
33

34. • Only if the slope and phase match
• Leveling mechanism
pushes the 2 parameterizations to their mid-value:
with
parameterization
arc-length
34

35. 35

36. 36

37. • Segments fitting
37

38. • Arcs fitting
38

39. 39

40. 40

41. 41

42. 42

43. 43

44. Input samples Active strokes
Robustness toward noise / outliers
 Temporal smoothing
44

45. • New propagation primitive
 for image space lines
 fully automatic
 running at interactive frame rates
• Temporally coherent basis for complex effects
 Arc-length variations (texture, offsets, tapering)
 Shape abstraction
45

46. Thank you for your attention
46